S2t streaming

[jakmro]

No description provided.

[Copilot]

Pull request overview

This PR introduces streaming transcription functionality for real-time audio processing using the Whisper model. The implementation uses a confirmation mechanism based on fuzzy matching to progressively confirm transcribed text as new audio chunks are processed.

Key Changes

• New streaming transcription API with four functions: init, insert, process, finalize, and destroy
• Fuzzy matching algorithm using Levenshtein distance to confirm stable transcription results
• Comprehensive documentation with usage examples for the streaming API

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 10 comments.

File	Description
cactus/ffi/cactus_ffi.h	Adds function declarations for the new streaming transcription API
cactus/ffi/cactus_stream_transcribe.cpp	Implements streaming transcription with audio buffering, fuzzy matching, and incremental confirmation logic
tests/test_engine.cpp	Adds test case for streaming transcription with chunked audio processing
docs/cactus_engine.md	Provides comprehensive documentation for all streaming API functions with examples and response formats

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.